[test] Add tests for previously failing interactive sessions

[dpiparo]

fixes #6439

[ferdymercury]

Thanks for the test!

There was another potential test case mentioned that could be added

size_t al(size_t p, size_t alignment) { return ((p + (alignment - 1)) & ~(alignment - 1)); } 
void printextra(size_t use, size_t p, size_t alignment) { auto realstart = al(p+use, alignment); std::cout << "alloc: " << p << " data: " << realstart << " waste: " << realstart - p - use << '\n' };
void printextra(size_t use, size_t p, size_t alignment) { auto realstart = al(p+use, alignment); std::cout << "alloc: " << p << " data: " << realstart << " waste: " << realstart - p - use << '\n'; };
printextra(24, 56, 64)

that lead to cling JIT session error: Failed to materialize symbols: { (main, { __ZNSt3__1lsB7v160006INS_11char_traitsIcEEEERNS_13basic_ostreamIcT_EES6_c }) }

which is also now solved in master.

[github-actions]

Test Results

    23 files      23 suites   3d 12h 56m 10s ⏱️
 3 876 tests  3 876 ✅ 0 💤 0 ❌
78 926 runs  78 926 ✅ 0 💤 0 ❌

Results for commit eb0f9b9.

♻️ This comment has been updated with latest results.

[dpiparo]

Thanks a lot @ferdymercury . Just to be sure: what is the expected output of that snippet?

[ferdymercury]

Just to be sure: what is the expected output of that snippet?

It's coming from #14788

but I believe the expected output should be:

ROOT_prompt_1:1:196: error: expected ';' after expression
alloc: 56 data: 128 waste: 48

instead of

cling JIT session error: Failed to materialize symbols: { (main, { __ZNSt3__1lsB7v160006INS_11char_traitsIcEEEERNS_13basic_ostreamIcT_EES6_c }) }

and or in macos

You are probably missing the definition of std::__1::basic_ostream<char, std::__1::char_traits<char>>& std::__1::operator<<[abi:ne190102]<std::__1::char_traits<char>>(std::__1::basic_ostream<char, std::__1::char_traits<char>>&, char)
Maybe you need to load the corresponding shared library?
IncrementalExecutor::executeFunction: symbol '_ZNKSt3__18io

[dpiparo]

Thanks. I wanted to make sure I got it right.
This is interesting: I see your producing the expected output on linux, however not on macos26.

IncrementalExecutor::executeFunction: symbol '_ZNKSt3__18ios_base5flagsB9nqe210106Ev' unresolved while linking [cling interface function]!
You are probably missing the definition of std::__1::ios_base::flags[abi:nqe210106]() const
Maybe you need to load the corresponding shared library?
IncrementalExecutor::executeFunction: symbol '_ZNSt3__1lsB9nqe210106INS_11char_traitsIcEEEERNS_13basic_ostreamIcT_EES6_c' unresolved while linking [cling interface function]!
You are probably missing the definition of std::__1::basic_ostream<char, std::__1::char_traits<char>>& std::__1::operator<<[abi:nqe210106]<std::__1::char_traits<char>>(std::__1::basic_ostream<char, std::__1::char_traits<char>>&, char)
Maybe you need to load the corresponding shared library?
IncrementalExecutor::executeFunction: symbol '_ZNSt3__116__pad_and_outputB9nqe210106IcNS_11char_traitsIcEEEENS_19ostreambuf_iteratorIT_T0_EES6_PKS4_S8_S8_RNS_8ios_baseES4_' unresolved while linking [cling interface function]!
You are probably missing the definition of std::__1::ostreambuf_iterator<char, std::__1::char_traits<char>> std::__1::__pad_and_output[abi:nqe210106]<char, std::__1::char_traits<char>>(std::__1::ostreambuf_iterator<char, std::__1::char_traits<char>>, char const*, char const*, char const*, std::__1::ios_base&, char)
Maybe you need to load the corresponding shared library?

This might potentially be a different problem.

[ferdymercury]

This might potentially be a different problem.

Ahh, then maybe we should reopen #14788 to decouple it from #6439 ?

Btw #19260 still fails on Linux and looks similarish to your mac26 failure

[dpiparo]

Ahh, then maybe we should reopen #14788 to decouple it from #6439 ?

That is a smart proposal: why not. Let's see how the tests go though before doing anything. Clearly we are dealing with some sort of platform dependent issues, and the battery of the CI could be of help to gather more information.

[dpiparo]

Windows deserves some special attention. For what concerns #6439, I think the CI pass proves that the problem is fixed. However, #7970 is not yet fixed. It's early for me to formulate a comprehensive analysis behind the failure.

[dpiparo]

The last change is an attempt to prove the point made above.